Diagnose wasted paid-ad spend, pacing, and allocation

Paid Ads Optimization

by nowork-studio · MIT license · ★ 3,853 on GitHub · updated Sep 24

Diagnose wasted paid-ad spend, pacing, and allocation, then propose safe evidence-backed optimizations. Use for waste, negatives, budgets, bid changes, poor CPA or ROAS, underpacing, overspend, or scaling decisions.

---
name: paid-ads-optimize
description: Diagnose wasted paid-ad spend, pacing, and allocation, then propose safe evidence-backed optimizations. Use for waste, negatives, budgets, bid changes, poor CPA or ROAS, underpacing, overspend, or scaling decisions.
argument-hint: "<campaign, target CPA/ROAS, or 'find waste'>"
---

# Paid Ads Optimization

Read `../shared/operating-contract.md` and `../shared/measurement-framework.md`. Review before changing anything.

## Diagnose before cutting

Verify the conversion signal, period completeness, spend volume, attribution model, and recent account changes. Spend with no recorded conversion can indicate broken tracking or immature data; treat it as a hypothesis until the signal and volume support an intervention. Check landing-page or operational failures before blaming targeting.

Classify the bottleneck as query/audience quality, creative fatigue, delivery/rank, budget constraint, landing-page mismatch, tracking, or economics. Use the specialized Google, Meta, X, LinkedIn, Reddit, or TikTok skill for live diagnosis. For other platforms, analyze only the supplied or verified data.

## Rank reversible moves

Overview

Read ../shared/operating-contract.md and ../shared/measurement-framework.md. Review before changing anything.

Manual install (npx, ZIP)

Claude Codenpx skills add nowork-studio/notfair-plugin --skill paid-ads-optimize -g -a claude-code

Claude Code (manual)

mkdir -p ~/.claude/skills && curl -sL https://codeload.github.com/nowork-studio/notfair-plugin/tar.gz/5229d2163795 \
  | tar -xz -C ~/.claude/skills --strip-components=2 notfair-plugin-5229d2163795/paid-ads/paid-ads-optimize

Windows (PowerShell)

New-Item -ItemType Directory -Force "$HOME\.claude\skills" | Out-Null; Invoke-WebRequest "https://m.agentalley.io/zip/92d046a61b13cee15766/paid-ads-optimize.zip" -OutFile "$env:TEMP\paid-ads-optimize.zip"; Expand-Archive "$env:TEMP\paid-ads-optimize.zip" -DestinationPath "$HOME\.claude\skills" -Force

Claude.ai / desktopDownload the ZIP and upload it under Settings → Capabilities → Skills.

Codex, Cursor, other agentsPoint them at SKILL.md. The other files are plain text.

Connects with Agent Alley MCP · agentalley.io/mcp

Files included

SKILL.md2.1 KB
agents/1 file · 220 B

Read the files here

nowork-studio/5229d212 files

SKILL.md2.1 KB

Shared references2SKILL.md reads these from elsewhere in the repo. Not in the ZIP.

Get them on GitHub ↗

SKILL.md

Show the full text26 lines

paid-ads-optimize/SKILL.md26 lines · 2.1 KB

PreviewCode

Outline

RawView on GitHub

Paid Ads Optimization

Read ../shared/operating-contract.md and ../shared/measurement-framework.md. Review before changing anything.

Diagnose before cutting

Verify the conversion signal, period completeness, spend volume, attribution model, and recent account changes. Spend with no recorded conversion can indicate broken tracking or immature data; treat it as a hypothesis until the signal and volume support an intervention. Check landing-page or operational failures before blaming targeting.

Classify the bottleneck as query/audience quality, creative fatigue, delivery/rank, budget constraint, landing-page mismatch, tracking, or economics. Use the specialized Google, Meta, X, LinkedIn, Reddit, or TikTok skill for live diagnosis. For other platforms, analyze only the supplied or verified data.

Rank reversible moves

Prefer this order: exclude an irrelevant query, placement, or audience; pause the narrowest losing unit; adjust budget or bid in a measured step; then consider structural change. For a reallocation, show the current and proposed allocations, the same total budget unless the user approves an increase, and the observable hypothesis.

Do not declare a loser from a few clicks. Set a threshold appropriate to the named target CPA, conversion lag, and channel role. Preserve upper-funnel and assisted-conversion context rather than judging all campaigns on last-click CPA alone.

Approval and follow-up

Present each exact mutation with scope, current value, proposed value, currency exposure, rationale, and review date. After approval, execute only through the verified platform skill or connector, read back the result, and record the intervention's expected effect and guardrail. Revisit after the declared observation window instead of promising a generic ongoing watch.

1	---
2	name: paid-ads-optimize
3	description: Diagnose wasted paid-ad spend, pacing, and allocation, then propose safe evidence-backed optimizations. Use for waste, negatives, budgets, bid changes, poor CPA or ROAS, underpacing, overspend, or scaling decisions.
4	argument-hint: "<campaign, target CPA/ROAS, or 'find waste'>"
5	---
6
7	# Paid Ads Optimization
8
9	Read `../shared/operating-contract.md` and `../shared/measurement-framework.md`. Review before changing anything.
10
11	## Diagnose before cutting
12
13	Verify the conversion signal, period completeness, spend volume, attribution model, and recent account changes. Spend with no recorded conversion can indicate broken tracking or immature data; treat it as a hypothesis until the signal and volume support an intervention. Check landing-page or operational failures before blaming targeting.
14
15	Classify the bottleneck as query/audience quality, creative fatigue, delivery/rank, budget constraint, landing-page mismatch, tracking, or economics. Use the specialized Google, Meta, X, LinkedIn, Reddit, or TikTok skill for live diagnosis. For other platforms, analyze only the supplied or verified data.
16
17	## Rank reversible moves
18
19	Prefer this order: exclude an irrelevant query, placement, or audience; pause the narrowest losing unit; adjust budget or bid in a measured step; then consider structural change. For a reallocation, show the current and proposed allocations, the same total budget unless the user approves an increase, and the observable hypothesis.
20
21	Do not declare a loser from a few clicks. Set a threshold appropriate to the named target CPA, conversion lag, and channel role. Preserve upper-funnel and assisted-conversion context rather than judging all campaigns on last-click CPA alone.
22
23	## Approval and follow-up
24
25	Present each exact mutation with scope, current value, proposed value, currency exposure, rationale, and review date. After approval, execute only through the verified platform skill or connector, read back the result, and record the intervention's expected effect and guardrail. Revisit after the declared observation window instead of promising a generic ongoing watch.
26